Presentation is loading. Please wait.

Presentation is loading. Please wait.

EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS

Similar presentations


Presentation on theme: "EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS"— Presentation transcript:

1 EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS
Lecture 3, 4/6/2005 University of Washington, Department of Electrical Engineering Spring 2005 Instructor: Professor Jeff A. Bilmes 4/6/2005 EE562

2 Today: Basic (Uninformed) Search
Chapter 3 4/6/2005 EE562

3 Reading & Homework For today: read chapter 3
For Monday: read chapter 4 Homework: Due Today: Next HW will be passed out either on Monday or will go out by (so stay tuned to ). 4/6/2005 EE562

4 Outline Problem-solving agents Problem types Problem formulation
Example problems Basic search algorithms 4/6/2005 EE562

5 Problem-solving agents
4/6/2005 EE562

6 Example: Romania On holiday in Romania; currently in Arad.
Flight leaves tomorrow from Bucharest Formulate goal: be in Bucharest Formulate problem: states: various cities actions: drive between cities Find solution: sequence of cities, e.g., Arad, Sibiu, Fagaras, Bucharest 4/6/2005 EE562

7 Example: Romania 4/6/2005 EE562

8 Problem types Deterministic, fully observable  single-state problem
Agent knows exactly which state it will be in; solution is a sequence Non-observable  sensorless problem (conformant problem) Agent may have no idea where it is; solution is a sequence Nondeterministic and/or partially observable  contingency problem percepts provide new information about current state often interleave search, execution Conditional (contingency) execution (based on current questions about current conditions) Unknown state space  exploration problem random walks/search 4/6/2005 EE562

9 Example: vacuum world Single-state, start in #5. Solution? 4/6/2005
EE562

10 Example: vacuum world Single-state, start in #5. Solution? [Right, Suck] Sensorless, start in {1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? 4/6/2005 EE562

11 Example: vacuum world Sensorless, start in {1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? [Right,Suck,Left,Suck] Contingency Nondeterministic: Suck may dirty a clean carpet Partially observable: location, dirt at current location. Percept: [L, Clean], i.e., start in #5 or #7 Solution? 4/6/2005 EE562

12 Example: vacuum world Sensorless, start in {1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? [Right,Suck,Left,Suck] Contingency Nondeterministic: Suck may dirty a clean carpet Partially observable: location, dirt at current location. Percept: [L, Clean], i.e., start in #5 or #7 Solution? [Right, if dirt then Suck] 4/6/2005 EE562

13 Single-state problem formulation
A problem is defined by four items: initial state e.g., "at Arad" actions or successor function S(x) = set of action–state pairs e.g., S(Arad) = {<Arad  Zerind, Zerind>, … } goal test, can be explicit, e.g., x = "at Bucharest" implicit, e.g., Checkmate(x) path cost (additive) e.g., sum of distances, number of actions executed, etc. c(x,a,y) is the step cost, assumed to be ≥ 0 A solution is a sequence of actions leading from the initial state to a goal state 4/6/2005 EE562

14 Selecting a state space
Real world is absurdly complex  state space must be abstracted for problem solving (Abstract) state = set of real states (Abstract) action = complex combination of real actions e.g., "Arad  Zerind" represents a complex set of possible routes, detours, rest stops, etc. For guaranteed realizability, any real state "in Arad“ must get to some real state "in Zerind" (Abstract) solution = set of real paths that are solutions in the real world Each abstract action should be "easier" than the original problem Hierarchical abstraction layers (divide and conquer) 4/6/2005 EE562

15 Vacuum world state space graph
actions? goal test? path cost? 4/6/2005 EE562

16 Vacuum world state space graph
states? integer dirt and robot location actions? Left, Right, Suck goal test? no dirt at all locations path cost? 1 per action 4/6/2005 EE562

17 Example: The 8-puzzle states? actions? goal test? path cost? 4/6/2005
EE562

18 Example: The 8-puzzle states? locations of tiles
actions? move blank left, right, up, down goal test? = goal state (given) path cost? 1 per move [Note: optimal solution of n-Puzzle family is NP-hard] 4/6/2005 EE562

19 Example: robotic assembly
states?: real-valued coordinates of robot joint angles parts of the object to be assembled actions?: continuous motions of robot joints goal test?: complete assembly path cost?: time to execute 4/6/2005 EE562

20 Tree search algorithms
Basic idea: offline, simulated exploration of state space by generating successors of already-explored states (a.k.a.~expanding states) 4/6/2005 EE562

21 Tree search example 4/6/2005 EE562

22 Tree search example 4/6/2005 EE562

23 Tree search example 4/6/2005 EE562

24 Implementation: general tree search
4/6/2005 EE562

25 Implementation: states vs. nodes
A state is a (representation of) a physical configuration A node is a data structure constituting part of a search tree includes state, parent node, action, path cost g(x), depth The Expand function creates new nodes, filling in the various fields and using the SuccessorFn of the problem to create the corresponding states. Q: How many and what would be the next states? 4/6/2005 EE562

26 Search strategies A search strategy is defined by picking the order of node expansion Strategies are evaluated along the following dimensions: completeness: does it always find a solution if one exists? time complexity: number of nodes generated space complexity: maximum number of nodes in memory optimality: does it always find a least-cost solution? Time and space complexity are measured in terms of b: maximum branching factor of the search tree d: depth of the least-cost solution m: maximum depth of the state space (may be ∞) 4/6/2005 EE562

27 Uninformed search strategies
Uninformed search strategies use only the information available in the problem definition Breadth-first search Uniform-cost search Depth-first search Depth-limited search Iterative deepening search 4/6/2005 EE562

28 Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end 4/6/2005 EE562

29 Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end 4/6/2005 EE562

30 Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end 4/6/2005 EE562

31 Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end 4/6/2005 EE562

32 Properties of breadth-first search
Complete? Yes (if b is finite) Time? 1+b+b2+b3+… +bd + b(bd-1) = O(bd+1) Space? O(bd+1) (keeps every node in memory) Optimal? Yes (if cost = 1 per step) Space is the bigger problem (more than time) 4/6/2005 EE562

33 Uniform-cost search Expand least-cost unexpanded node Implementation:
fringe = queue ordered by path cost (priority queue) Equivalent to breadth-first if step costs all equal Complete? Yes, if step cost ≥ ε Time? # of nodes with g ≤ cost of optimal solution, O(bceiling(C*/ ε)) where C* is the cost of the optimal solution Space? # of nodes with g ≤ cost of optimal solution, O(bceiling(C*/ ε)) Optimal? Yes – nodes expanded in increasing order of g(n) 4/6/2005 EE562

34 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

35 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

36 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

37 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

38 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

39 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

40 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

41 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

42 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

43 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

44 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

45 Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front 4/6/2005 EE562

46 Properties of depth-first search
Complete? No: fails in infinite-depth spaces, spaces with loops Modify to avoid repeated states along path  complete in finite spaces Time? O(bm): terrible if m is much larger than d but if solutions are dense, may be much faster than breadth-first Space? O(bm), i.e., linear space! Optimal? No 4/6/2005 EE562

47 Depth-limited search = depth-first search with depth limit l,
i.e., nodes at depth l have no successors Recursive implementation: 4/6/2005 EE562

48 Iterative deepening search
4/6/2005 EE562

49 Iterative deepening search l =0
4/6/2005 EE562

50 Iterative deepening search l =1
4/6/2005 EE562

51 Iterative deepening search l =2
4/6/2005 EE562

52 Iterative deepening search l =3
4/6/2005 EE562

53 Iterative deepening search
Number of nodes generated in a depth-limited search to depth d with branching factor b: NDLS = b0 + b1 + b2 + … + bd-2 + bd-1 + bd Number of nodes generated in an iterative deepening search to depth d with branching factor b: NIDS = (d+1)b0 + d b^1 + (d-1)b^2 + … + 3bd-2 +2bd-1 + 1bd For b = 10, d = 5, NDLS = , , ,000 = 111,111 NIDS = , , ,000 = 123,456 Overhead = (123, ,111)/111,111 = 11% 4/6/2005 EE562

54 Properties of iterative deepening search
Complete? Yes Time? (d+1)b0 + d b1 + (d-1)b2 + … + bd = O(bd) Space? O(bd) Optimal? Yes, if step cost = 1 4/6/2005 EE562

55 Summary of algorithms 4/6/2005 EE562

56 Repeated states Failure to detect repeated states can turn a linear problem into an exponential one! 4/6/2005 EE562

57 Graph search 4/6/2005 EE562

58 Other search Bi-directional search Searching with partial information
generate from both beginning and end, combine when the solutions meet can significantly reduce search space sometimes difficult to generate backwardss Searching with partial information a state in this case might be a set of “possible” states that we are in at a given time. Take actions to limit the set of states that we might be in (uncertainty reduction might also be a goal here). 4/6/2005 EE562

59 Example: vacuum world Sensorless, start in {1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? [Right,Suck,Left,Suck] 4/6/2005 EE562

60 Summary Problem formulation usually requires abstracting away real-world details to define a state space that can feasibly be explored Variety of uninformed search strategies Iterative deepening search uses only linear space and not much more time than other uninformed algorithms 4/6/2005 EE562


Download ppt "EE562 ARTIFICIAL INTELLIGENCE FOR ENGINEERS"

Similar presentations


Ads by Google